Nearly 75% of patients with CF have impaired glucose tolerance and 15% have frank diabetes. This is caused by a combination of insulin resistance and insulin deficiency. The 1998 CF Foundation Consortium recognized 4 major categories of glucose tolerance in CF: 1) Normal glucose tolerance (FBG <126; pp <150); 2) Impaired glucose tolerance (FBG <126; pp 151-200); 3) Diabetes w/o fasting hyperglycemia (FBG <126; pp >200), a type of impaired glucose tolerance. 4) Diabetes with fasting hyperglycemia. Only patients who have diabetes with fasting hyperglycemia routinely receive insulin treatment. Impaired glucose tolerance is associated with weight loss and possibly with increased morbidity in CF, yet insulin is currently not routinely used for management. The most common reason for lack of use of insulin is that many of these patients have normal FBG, despite very high post-prandial (pp) readings. Many centers only check fasting blood sugars and thus do not recognize impaired glucose tolerance. Furthermore, the most effective insulin regimen is multi-insulin injections around the time of meals (generally 3-5 shots per day). Many CF patients have been reluctant to take this large number of insulin injections. The insulin pump is connected to the patient via a small plastic catheter. The catheter is inserted with a 24 gauge needle (in a fashion similar to an IV catheter) just under the skin. The catheter is then changed every three days. The patient wearing the pump may then give himself or herself a ""bolus dose"" of insulin every time he or she eats, without needing to give a shot. A further advantage of the pump is that a very low dose of basal insulin may be given throughout the day. This low dose of insulin mimics the normal pancreas and may be especially advantageous for the insulin deficient CF patient. We will recruit 8 CF adults with impaired glucose tolerance to participate in this study. Each subject will be provided with the pump and with all materials needed for use with the pump. Each patient will also be provided with a glucose meter and test strips. He/she will be asked to wear the pump for three months and to check his/her blood sugar levels four times per day. Prior to the pump placement and at the end of three months, each patient will undergo the following measurements: 1) whole body protein turnover using the stable isotope [1-13C] leucine; 2) DEXA scan for measurement of lean body mass; 3) anthropometric measurements; 4) Hemoglobin A1c. Expected endpoints of the study include: weight gain, gain of lean body mass, increased protein synthesis and decreased protein breakdown.